The Venture Capital Podcast with Fexingo · 2026-08-19 · 7 min
In episode 162 of The Venture Capital Podcast, Lucas and Luna explore how venture capitalists are rethinking startup valuations as AI compute costs become the dominant line item on many young companies' profit-and-loss statements. Anchored by the latest TechCrunch story on a startup that helps Wall Street price AI compute, the conversation zeroes in on a single, concrete shift: the rise of compute-as-a-service agreements that tie usage fees to model performance and revenue milestones. Lucas breaks down the mechanics of these deals, using NVIDIA's recent five-day slide and the broader tech sell-off, including Shopify's seven-point-six percent drop, to illustrate why investors are suddenly demanding unit economics on compute the way they once did for cloud spend. Luna challenges the model with a founder's perspective, asking how a seed-stage startup is supposed to forecast its compute needs when model architectures are changing so fast. They also touch on the geopolitical angle - Chinese hackers and export controls - and what that means for supply.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.
How VCs Are Pricing Startup Burn Rate in 202652 / 100Why VCs Are Underwriting Startup Supply Chain ResilienceHow VCs Are Pricing Startup Exits in 2026How VCs Are Valuing Startup Community in 2026How VCs Are Pricing Startup Adaptability in 2026
